Abstract:
Providing credit has become a main activity for financial and non-financial institutions. Credit defined as an agreement for lending money and the borrower obligated to repay instalment with given interest. However, this transaction might lead into risk called credit risk. This risk occurred if borrowers incapable to comply with their obligations that generate loss for creditors. This research aim to form model that capable in distinguish between eligible and non-eligible prospective customers. This model later predicts the loan status will become default or fully paid. The method developed in this research is Backpropagation Neural Networks with applied activation function is a sigmoid function. Furthermore, this research will have three data types for analyzed, with first data type is every variable given dataset and for the rest data type is only variables with correlation with target variables. The results of this research show that within the data type, the highest accuracy of prediction is 94.37% while the lowest accuracy prediction is 80.28%. From this result, researcher concluded that Backpropagation Neural Network is excellent in predicting loan status of applicant’s.